Effect of Turosteride, a 5a-Reductase Inhibitor,on the Dunning R3327 Rat Prostatic Carcinoma

T. Zaccheo,* D. Giudici, and E. di Salle

Department of Endocrinology, R&D Oncology, Pharmacia, Nerviano (MI), Italy

BACKGROUND. Turosteride (FCE 26073) is a new, potent, and selective 5a-reductaseinhibitor. We have investigated its effect on tumor growth, organ weight, and serum hor-mone levels in rats bearing the androgen sensitive Dunning R3327 prostatic carcinoma.METHODS. Animals with tumor diameters of 0.5–1.5 cm were treated for 9 weeks withturosteride (50 and 200 mg/kg/day, 6 days a week, orally), flutamide (25 mg/kg/day, 6 daysa week, orally), and leuprolide (300 mg/rat, every 3 weeks, subcutaneously) or they werecastrated.RESULTS. Turosteride was ineffective at the dose of 50 mg/kg/day, whereas at 200 mg/kg/day it significantly decreased tumor growth by 45%. Flutamide and leuprolide were highlyeffective in reducing tumor growth (70 and 77%), although their effect was slightly lowerthan that of castration (85%). A significant reduction of ventral prostate weight occurred inrats treated with turosteride at 50 and 200 mg/kg/day (53% and 60%). In contrast to leupro-lide and castration, the inhibitory effect of turosteride on tumor growth and prostate weightwas not associated to any decrease in serum testosterone.CONCLUSIONS. Turosteride has antitumor activity on Dunning prostatic tumor growthand its role in prostatic cancer treatment is worthy of further investigation. Prostate 30:85–91,1997. © 1997 Wiley-Liss, Inc.

KEY WORDS: FCE 26073; flutamide; leuprolide; androgen levels

INTRODUCTION

The Dunning R3327 rat prostatic carcinoma is awell-characterized model system. It shares severalcharacteristics with human prostatic cancer includingslow growth rate, responsiveness to hormonal ther-apy, and histology similar to that of a well-differen-tiated carcinoma [1]. Moreover, it contains 5a-reduc-tase, the enzyme that catalyzes the metabolictransformation of testosterone (T) to 5a-dihydrotest-osterone (DHT), the androgen essential for prostategrowth and function [2].

Turosteride (FCE 26073; 1-[4-methyl-3-oxo-4-aza-5a-androstane-17b-carbonyl]-1, 3-diisopropylurea) isa new, potent, and selective 5a-reductase inhibitor[3]. When tested on rat and human prostatic 5a-re-ductases, the compound showed a similar degree ofpotency with IC50 values of 53 and 55 nM, respec-tively [3,4]. Twenty-day oral administration of turo-steride to adult male rats at doses of 3, 10, or 30 mg/kg/day was found to cause a reduction in prostateweight associated with a reduction of prostatic DHT

content by 61, 74, and 78%, respectively, without anyrelevant increase in prostatic T content [5].

In this study we have investigated the effect ofturosteride on the Dunning R3327 rat prostatic tumorgrowth in comparison with the effect of other andro-gen withdrawal therapies, i.e., the androgen receptorantagonist flutamide, the luteinizing hormone-re-leasing hormone (LH-RH) agonist leuprolide, or cas-tration. In addition, the effect of these compounds onendocrine organ weights and serum hormone levelswas also investigated.

*Correspondence to: Dr. Tiziana Zaccheo, Department of Endocri-nology, Pharmacia, Via Giovanni XXIII 23, 20014 Nerviano (MI),Italy.Received 5 June 1995; Accepted 8 January 1996

The Prostate 30:85–91 (1997)

© 1997 Wiley-Liss, Inc.

MATERIALS AND METHODS

Animals

Male Copenhagen rats, weighing approximately200 g, were supplied by Harlan Sprague Dawley Inc.(IN, USA) and imported by Nossan s.r.l. (Correz-zana, Italy). Animals were housed in temperature-controlled rooms (22 ± 2°C) on a circadian rhythm of12 hr of light (6 AM to 6 PM) and 12 hr of darkness.

Prostatic Tumor Model

The Dunning R3327 prostatic carcinoma waskindly provided by Dr. H. Altman, Papanicolau Can-cer Research Center, University of Miami (FL, USA).The tumor was maintained by serial transplantationin Copenhagen rats. It was passed by harvestingfresh tumor, dicing it with sterile scissors in a sterile0.9% NaCl solution, and aseptically implanting a sin-gle tumor fragment (approximately 3–4 mm in size)subcutaneously in the flank of the recipient animalunder mild diethyl-ether anesthesia.

Treatment and Experimental Set-Up

Tumor-bearing animals with tumor diameters of0.5–1.5 cm were randomly assigned to the differenttreatment groups. Turosteride (synthesized in thePharmacia Research Laboratories, Italy) and fluta-mide (supplied by Schering-Plough, Italy) were sus-pended in 0.5% Methocel (methylcellulose 400, DowChemical, Midland, MI) containing 0.4% Tween 80(Merck, Darmstadt, Germany) and administeredorally, 6 days a week for 9 weeks. Leuprolide (Enan-tone depott, Takeda Italia Farmaceutici, Catania, It-aly) was given by subcutaneous injection of the depotformulation every 3 weeks, on days 0 (first experi-mental day), 21, and 42. The compounds were givenin a volume of 5 ml/kg. Castration was performed onday 0, under diethyl-ether anesthesia.

Tumor growth was followed by measuring the twoperpendicular diameters with calipers, and tumorweight was calculated according to the formula: d2 ×D/2, where d is the minimum and D is the maximumdiameter [6]. The weight of each tumor was expressed as a percentage of initial tumor weight, mea-sured on day 0, and taken as 100%. The area underthe percentage of initial tumor weight vs. time curvewas calculated by the linear trapezoidal method forup to 9 weeks (AUC 0–9 weeks). Animals were sac-rificed at the end of the 9th week (24 hr after the lastdose for the groups treated with turosteride or fluta-mide) and blood samples were collected. Serum ob-tained after centrifugation was stored for hormonaldeterminations at −20°C. In addition, the following

organs were removed and weighed: the ventral pros-tate, seminal vesicles, testes, and anterior pituitary.

In the first experiment five treatment groups (n =8–10) were included: control group (treated orallywith the vehicle), turosteride (50 mg/kg/day), fluta-mide (25 mg/kg/day), leuprolide (300 mg per injectionper rat), and castrated group (treated orally with thevehicle). In the second experiment the following fivegroups (n = 6–9) were included: control group, tur-osteride (200 mg/kg/day), leuprolide (300 mg per in-jection per rat), turosteride combined with leupro-lide, and castration.

Serum Hormone Assays

Serum concentrations of T, DHT, and androstene-dione (A) were measured by specific radioimmunoas-says (RIA) after sample extraction and extract purifi-cation on a celite column. The serum sample (1.5 ml)was extracted twice with 6 ml diethyl-ether. Thedried ether phase was dissolved in 0.5 ml isooctaneand applied to a celite column (1.2 g of ChromatolitheAt, supplied by bioMerieux (Marcy l’Etoile, France),in 5 ml glass pipette). Elution was then performedwith increasing concentrations of ethylacetate inisooctane (0, 6, and 22%). The fractions containing A(0%), DHT (6%), and T (22%) were evaporated underN2 and the dried samples were used for RIA. A,DHT, and T levels in the suspended samples wereestimated, in duplicate, by using the [3H]A and [3H]TRIA kits supplied by bioMerieux (France) and the[3H]DHT kit supplied by ICN Biomedicals (CA,USA). The final sensitivity for A assay was 0.045 ng/ml, for T assay it was 0.066 ng/ml, and for DHT assayit was 0.05 ng/ml of serum.

Serum LH was measured by a direct assay using adouble antibody RIA, supplied by Amersham (Inter-national plc., Little Chalfont, UK). The sensitivity ofthe assay was 0.2 ng/ml of serum.

RESULTS

Effect on Tumor Growth

Figure 1 and Table I show the effect of turosteride(50 mg/kg/day orally, 6 days a week), flutamide (25mg/kg/day orally, 6 days a week), and leuprolide (300mg per rat subcutaneously, every 3 weeks) given for 9weeks. In addition, one group of animals was cas-trated the first experimental day to check the hor-mone sensitivity of the tumor. The weekly recordedtumor weights (expressed as percentage of the initialweight) are reported in Figure 1. The AUC (0–9week) values (expressed as percent × week) are re-ported in Table I. During the 9-week observation pe-riod the Dunning R3327 tumor grew progressively in

86 Turosteride on Rat Prostatic Tumor

the control group (AUC, mean ± SE = 17,079 ±4,216), whereas castration resulted in a marked de-crease in tumor growth (2,526 ± 559), demonstratingthe androgen responsiveness of the tumor. Treat-ment with turosteride had no effect on tumor growth(14,773 ± 3,875). Flutamide and leuprolide werehighly effective in reducing tumor growth (5,053 ±1,448 and 3,902 ± 833, respectively), although theireffect was slightly lower than that of castration.

In a further experiment the effect of a 9-week treat-ment with turosteride at a higher dose level (200 mg/kg/day orally) was investigated. In addition, the ef-fect of the combined treatment with turosteride andleuprolide was also investigated. Results are illus-trated in Figure 2 and Table II. Turosteride, at thedose of 200 mg/kg/day, significantly decreased tumorgrowth (AUC 7,082 ± 1,441, vs. 12,821 ± 1,097 ofcontrols). Treatment with leuprolide resulted in ahigher suppression of tumor growth (3,169 ± 324).Combined treatment with turosteride and leuprolidedid not result in a better antitumor effect (5,044 ±1,113) in comparison with leuprolide alone. Castra-tion produced a more remarkable reduction of tumorgrowth (1,829 ± 358) than all other treatments.

Effect on Organ Weights

The ventral prostate, seminal vesicles, testes, andanterior pituitary of tumor-bearing rats were excisedand weighed at the 9-week sacrifice. Since the differ-ent treatments slightly reduced the body weight gain(more pronounced for flutamide), the relativeweights of these organs are reported in Figures 3 and

4. Turosteride at 50 mg/kg/day orally caused a de-crease of ventral prostate and seminal vesicle weightof 53% and 37%, respectively (Fig. 3). Flutamidecaused a decrease of ventral prostate (76%) and ofseminal vesicle (52%) weight, and a slight increase(19%) of testis weight. Leuprolide caused a decreaseof ventral prostate (82%), seminal vesicle (68%), andtestis (50%) weight. Castration, similar to the effect ofleuprolide, reduced ventral prostate (88%) and sem-inal vesicle (62%) weight.

In the second experiment (Fig. 4) turosteride, atthe dose of 200 mg/kg/day orally, reduced prostateweight by 60% and seminal vesicles by 35%; no effecton testis weight was observed. The inhibitory effectsof leuprolide and castration were similar, as in theprevious experiment. Combined treatment with tur-osteride and leuprolide caused a reduction of prostateand seminal vesicle (83% and 66%) weights similar tothat caused by leuprolide alone (84% and 67%). Testisweight was reduced by 76% with the combined treat-ment and by 62% with leuprolide alone. No effect onpituitary weight was observed with any treatmentmodality.

Effect on Serum Hormone Levels

Turosteride, at the dose of 50 mg/kg/day, slightlyreduced serum DHT level without affecting serum T,A, and LH levels (Fig. 5). Flutamide caused a markedincrease in serum DHT, T, A, and LH levels. Leupro-lide slightly reduced DHT and considerably reducedT, A, and LH levels. Castration did not reduce DHTlevel, while T and A were strongly reduced; LH wasincreased several times, as expected.

In the second experiment, turosteride, at the dose of200 mg/kg/day orally, slightly reduced serum DHTlevel, as did castration (Fig. 6). The compound didnot affect serum T level and only slightly increasedthe A level. The combined treatment with leuprolideand turosteride caused a slight reduction of DHT andLH levels and a relevant reduction of T and A levels.

DISCUSSION

The results presented indicate that the 5a-reduc-tase inhibitor turosteride is effective in reducing theDunning prostatic tumor growth. In fact the com-pound, at the oral dose of 200 mg/kg/day given for 9weeks, caused a reduction of tumor growth by 45%,associated to a reduction of ventral prostate weightby 60%. Both these effects, however, were found tobe less marked than those observed with the antian-drogen flutamide and the LH-RH agonist leuprolide,whose effects were similar to those of castration.

The endocrine treatment of prostate cancer has tra-ditionally consisted of surgical or chemical castration

Fig. 1. Effect of 9-week treatment with turosteride, flutamide,or leuprolide on the growth of Dunning prostatic carcinoma in rats(8–10 animals per group). Turosteride and flutamide were givenorally 6 days a week; leuprolide was given subcutaneously every 3weeks. Treatments started when mean tumor diameters were 0.5–1.5 cm. Tumor growth was expressed as percentage of the initialtumor weight.

Zaccheo et al. 87

in order to lower serum T. However, such a systemicreduction in the circulating T level has several sideeffects including sterility, impotence, decreased li-bido, hot flashes, and loss of anabolic effects of an-drogens on muscle mass [7,8]. Recently a new class ofdrugs has been developed that inhibits the conver-sion of T to the more active androgen DHT, throughinhibition of the prostatic 5a-reductase enzyme [9].Treatment with such inhibitors results, both in exper-imental animals and in humans, in lowering prostaticDHT and consequently in involution of the prostate

size. 5a-reductase inhibitor treatment does not resultin a reduction of serum T and therefore does not in-duce the side effects observed with antiandrogens[10]. The first 5a-reductase inhibitor developed, fin-asteride, has recently been introduced into the mar-ket for the therapy of symptomatic benign prostatichyperplasia (BPH) [10,11].

The value of 5a-reductase inhibitors in the therapyof prostate cancer remains to be determined. Con-flicting results have been reported about the experi-mental antitumor activity of these compounds.Brooks et al. [12] have demonstrated that finasteride,given at the oral dose of 25 mg/kg/day, was inactive inthe Dunning R3327 rat prostatic cancer, whereas theventral prostate of treated rats was reduced by 64%.In the subline H of the Dunning tumor both finas-teride and epristeride, at the oral dose of 100 mg/kg/day, have been reported to be inactive by Lamb et al.[13]. The same authors demonstrated that epris-teride, at doses of 50 and 100 mg/kg/day, has amarked antitumor activity on the G subline of theDunning tumor, a fast growing, poorly differentiatedsubline with low levels of 5a-reductase activity. Un-fortunately, they have not studied the effect of finas-teride on such a subline. Since the reduction in tissueDHT, secondary to 5a-reductase inhibition, appearsto be more difficult to achieve in a prostatic tumorthan in normal prostate [13], it was suggested thatonly prostatic cancer with low 5a-reductase activity(like the G subline) could respond to 5a-reductaseinhibitors. Furthermore, the role of T, in addition tothat of DHT, in supporting prostatic cancer growthhas not been completely clarified and it has been hy-pothesized that androgenic stimulus to prostatic can-

TABLE I. Effect of 9-Week Treatment with Turosteride,Flutamide, Leuprolide, or Castration on Dunning

Prostatic Tumor Growth in Rats

Treatmentgroup

Noof

rats

AUC(0–9 weeks)(% × week)a

Inhibition(%)

Control 9 17,079 ± 4,216b

Turosteride50 mg/kg/day orally

9 14,773 ± 3,875 13.5

Flutamide25 mg/kg/day orally

10 5,053 ± 1,448* 70.3

Leuprolide 300 mg/ratsubcutaneously(every 3 weeks)

8 3,902 ± 833* 77.2

Castration 9 2,526 ± 599* 85.2

aAUC = area under the percent of initial tumor weight vs. timecurve, calculated by the linear trapezoidal method up to 9 weeks.bMean ± SE.*P < 0.01 vs. control group (Dunnett’s test).

TABLE II. Effect of 9-Week Treatment With Turosterideand Leuprolide Alone or in Combination on Dunning

Prostatic Tumor Growth in Rats

Treatmentgroup

No.of

rats

AUC(0–9 weeks)(% x week)a

Inhibition(%)

Control 9 12,821 ± 1,097b

Turosteride200 mg/kg/day orally

6 7,082 ± 1,441* 44.8

Leuprolide 300 mg/ratsubcutaneously(every 3 weeks)

9 3,169 ± 324** 75.3

Turosteride +Leuprolide

8 5,044 ± 1,113** 60.7

Castration 8 1,829 ± 538** 85.7

aAUC = area under the percent of initial tumor weight vs. timecurve, calculated by the linear trapezoidal method up to 9 weeks.bMean ± SE.*P < 0.05; **P < 0.01 vs. control group (Dunnett’s test).

Fig. 2. Effect of 9-week treatment with turosteride and leupro-lide, given alone or in combination, on the growth of Dunningprostatic carcinoma in rats (six to nine animals per group). Turo-steride was given orally 6 days a week; leuprolide was given sub-cutaneously every 3 weeks. Treatments started when mean tumordiameters were 0.5–1.5 cm. Tumor growth was expressed as per-centage of the initial tumor weight.

88 Turosteride on Rat Prostatic Tumor

cer could be provided by either DHT or T [11]. In thisrespect the observation that in rats turosteride is ableto inhibit prostatic DHT levels without causing a sec-ondary increase in tissue T concentration [5] couldexplain its antitumor activity observed in the Dun-ning model not sensitive to finasteride.

In the present study when turosteride was admin-istered at the lower oral dose of 50 mg/kg/day, noeffect on tumor growth was observed, whereas pros-tate weight was inhibited by 53%. This suggests a

different sensitivity of normal prostate and prostatictumor to 5a-reductase inhibitor therapy. In fact, theinhibitory effect on the tumor (45%) was observed at200 mg/kg/day, a dose level causing only a slight fur-ther increase in the inhibitory effect on prostateweight (from 53% to 60%). As also shown here, 5a-reductase inhibitors have been reported to cause lessinvolution of the rat ventral prostate than does cas-tration [14,15]; in fact they were found not to stimu-late apoptosis and to induce a less marked reduction

Fig. 3. Effect of 9-week treatment with turosteride (50 mg/kg/day orally), flutamide (25 mg/kg/day orally), or leuprolide (300µg/rat subcutaneously, every 3 weeks) on the relative organ

weights of rats bearing Dunning prostatic carcinoma. Bars repre-sent mean ± SE (8–10 animals per group). **P < 0.01 vs. controlsby Dunnett’s test.

Fig. 4. Effect of 9-week treatment with turosteride (200 mg/kg/day orally) and leuprolide (300 µg/rat subcutaneously, every 3weeks) given alone or in combination, on the relative organ

weights of rats bearing Dunning prostatic carcinoma. Bars repre-sent mean ± SE (six to nine animals per group). **P < 0.01 vs.controls by Dunnett’s test.

Fig. 5. Effect of 9-week treatment with turosteride (50 mg/kg/day orally), flutamide (25 mg/kg/day orally), or leuprolide (300µg/rat subcutaneously, every 3 weeks) on serum DHT, T, A, andLH levels in rats bearing Dunning prostatic carcinoma. Rats were

sacrificed 24 hr after the last oral dose. Bars represent mean ± SE(8–10 animals per group). **P < 0.01 vs. controls by Dunnett’stest.

Zaccheo et al. 89

in DNA synthesis in comparison with castration[16,17]. Incomplete inhibition of DHT synthesis ismost likely the cause for such lower effects. In addi-tion, the increased prostatic concentration of T, usu-ally related to 5a-reductase inhibition, could preventgreater involution of the prostate size and greater lossof cells. Similarly, or perhaps more than in the nor-mal prostate, T levels in tumor tissue could partlymaintain tumor growth when DHT synthesis is in-hibited.

Molecular cloning studies have identified twogenes that encode the type 1 and type 2 isozymes of5a-reductase [18]. It has been recently shown that inhuman prostatic cancer the 5a-reductase enzyme isexpressed as the type 2 isoform in the stromal but notin the epithelial cells [19], which contain the andro-gen receptor [20]. Therefore, DHT acting on tumorcells must be derived from adjacent stroma (as is thecase of local tumor) or via the peripheral circulation(as is the case of metastatic lesions). The presence ofthe enzyme in tumor stroma, whose relative amountgradually decreases when tumor progresses, sug-gests that early-stage human prostate cancer could besensitive to DHT deprivation. In addition, the lowtoxicity and the absence of side effects render 5a-reductase inhibitors an interesting class of com-pounds to be tested for the possible preventive effectin the development of prostate cancer. A study withfinasteride in the prevention of human prostate can-cer has already been initiated [21].

REFERENCES

1. Isaacs JT, Coffey DS: Adaptation versus selection as themechanism responsible for the relapse of prostatic can-cer to androgen ablation therapy as studied in the Dun-ning R3327-H adenocarcinoma. Cancer Res 41:5070–5075, 1981.

2. Bruchovsky N, Rennie PS, Batzold FH, Goldenberg SL,

Fletcher T, Mc Loughlin MG: Kinetic parameters of 5a-reductase activity in stroma and epithelium of normal,hyperplastic, and carcinomatous human prostates. JClin Endocrinol Metab 67:806–816, 1988.

3. di Salle E, Briatico G, Giudici D, Ornati G, Nesi M,Panzeri A: 17b-Acylurea derivatives of 4-azasteroids asinhibitors of testosterone 5a-reductase. J Steroid Bio-chem Mol Biol 41:765–768, 1992.

4. di Salle E, Briatico G, Giudici D, Ornati G, Panzeri A:Endocrine properties of the testosterone 5a-reductaseinhibitor turosteride (FCE 26073). J Steroid BiochemMol Biol 48:241–248, 1994.

5. di Salle E, Giudici D, Briatico G, Ornati G, Panzeri A:Hormonal effects of turosteride, a 5a-reductase inhibi-tor, in the rat. J Steroid Biochem Mol Biol 46:549–555,1993.

6. Geran RI, Greenberg NH, Mc Donald MM, Schuma-cher AM, Abbott BJ: Protocol for screening chemicalagents and natural products against animal tumors andother biological systems. Cancer Chemother Rep 3:1–103, 1972.

7. Vogelzang NJ, Kennealey GT: Recent developments inendocrine treatment of prostate cancer. Cancer 70:966–976, 1992.

8. Hsieh W-S, Simons JW: Systemic therapy of prostatecancer. New concepts from prostate cancer tumor biol-ogy. Cancer Treatment Rev 19:229–260, 1993.

9. Brooks JR, Berman C, Primka RL, Reynolds GF, Ras-musson GH: 5a-Reductase inhibitory and antiandro-genic activities of some 4-azasteroids in the rat. Steroids47:1–19, 1986.

10. Gormley GJ, Stoner E, Bruskewitz RC, Imperato-McGinley J, Walsh PC, McConnell JD, Andriole GL,Geller J, Brachen BR, Tenover JS, Vaughan ED, PappasF, Taylor A, Binkowitz B, Ng J: The effect of finasteridein men with benign prostatic hyperplasia. N Engl J Med327:1185–1191, 1992.

11. Stoner E: The clinical development of a 5a-reductaseinhibitor, finasteride. J Steroid Biochem Mol Biol 37:375–378, 1990.

12. Brooks JR, Berman C, Nguyen H, Prahalada S, PrimkaRL, Rasmusson GH, Slater EE: Effect of castration,DES, flutamide, and the 5a-reductase inhibitor, MK-906, on the growth of the Dunning rat prostatic carci-noma, R3327. Prostate 18:215–227, 1991.

Fig. 6. Effect of 9-week treatment with turosteride (200 mg/kg/day orally) and leuprolide (300 µg/rat subcutaneously, every 3weeks) given alone or in combination, on serum DHT, T, A, and LHlevels in rats bearing Dunning prostatic carcinoma. Animals were

sacrificed 24 hr after the last oral dose. Bars represent mean ± SE(six to nine animals per group). *P < 0.05; **P < 0.01 vs. controlsby Dunnett’s test.

90 Turosteride on Rat Prostatic Tumor

13. Lamb JC, Levy MA, Johnson RK, Isaacs JT: Response ofrat and human prostatic cancers to the novel 5a-reduc-tase inhibitor, SKF 105657. Prostate 21:15–34, 1992.

14. Ghusn H, Shao TC, Klima M, Cunninghamm GR:4-MAPC, a 5a-reductase inhibitor, reduces rat ventralprostate weight, DNA and prostatein concentrations. JAndrol 12:315–322, 1991.

15. Rittmaster RS, Magor KE, Manning AP, Norman RW,Lazier CB: Differential effect of 5a-reductase inhibitionand castration on androgen regulated gene expressionin rat prostate. Mol Endocrinol 5:1023–1029, 1991.

16. Shao TC, Kong A, Cunningham GR: Effects of finas-teride on the rat ventral prostate. J Androl 14:79–86,1993.

17. Isaacs JT, Lundmo PI, Berges R, Martikainen P, Ky-prianou N, English HF: Androgen regulation of pro-

grammed cell death of normal and malignant prostaticcells. J Androl 13:457–464, 1992.

18. Andersson S, Russell DW: Structural and biochemicalproperties of cloned and expressed human and rat ste-roid 5a-reductases. Proc Natl Acad Sci USA 87:3640–3644, 1990.

19. Silver RI, Wiley EL, Davis DL, Thigpen AE, RussellDW, McConnell JD: Expression and regulation of ste-roid 5a-reductase 2 in prostate disease. J Urol 152:433–437, 1994.

20. Liao SS, Kokontis J, Sai T, Hiipakka RA: Androgen re-ceptors: Structures, mutations, antibodies and cellulardynamics. J Steroid Biochem 34:41–45, 1989.

21. Ford LG, Brawley OW, Perlman JA, Nayfield SG,Johnson KA, Kramer BS: The potential for hormonalprevention trials. Cancer 74:2726–2733, 1994.

Zaccheo et al. 91